Skip to main content
. 2018 Dec 4;9:744. doi: 10.3389/fendo.2018.00744

Table 3.

Main findings from T2D studies investigating genome-wide DNA methylation in human population-based studies.

Author (year) Population Sample size Gender Tissue type Method Study outcome
Toperoff et al. (79) Jewish T2D = 710 Controls = 459 M and F WB Microarray-based methylation assays Differential methylation identified in 13 CpGs, mapping to SLC30A8, TCF7L2, KCNQ1, FTO, THADA, and JAZF1 genes in T2D subjects compared to controls.
Chambers et al. (80) Indian Asian and European Indian Asian: T2D = 1,608 Controls = 11 927 European: T2D = 306 Controls = 6,760 M and F PB 450 K Differential methylation identified in five regions mapping to TCF7L2, FTO, KCNQ1, TXNIP ABCG1, PHOSPHO1, SOCS3, and SREBF1 genes, replicated in two cohorts.
Dayeh et al. (81) European T2D = 19 Controls = 19 M and F WB 450 K ABCG1, PHOSPHO1 associated with future T2D risk but not SOCS3, SREBF1 or TXNIP. ABCG1 hypermethylation positively associated with HbA1c and fasting insulin levels.
Kriebel et al. (82) German 1,448 non-diabetic (FBG and HbA1c) 1,440 non-diabetic (FI and HOMA-IR) 617 non-diabetic (2-h insulin) M and F WB 450 K DNA methylation at cg06500161 (ABCG1) associated with fasting glucose, fasting insulin, and HOMA-IR.
Hidalgo et al. (83) American aHealthy individuals = 544 bHealthy individuals = 293 M and F WB 450 K ABCG1 hypermethylation associated with fasting insulin and HOMA-IR.
Walaszczyk et al. (84) Dutch T2D = 100 Controls = 100 M and F WB 450 K Differential methylation of ABCG1, LOXL2, TXNIP, SLC1A5, and SREBF1 associated with T2D.
Muftah et al. (19) aArab, bCaucasian aT2D = 30 Controls = 93 b180 twins from TwinsUK cohort M and F WB 450 K Differential methylation identified in TXNIP and DQX1 genes in T2D subjects compared to controls.
Kulkarni et al. (85) Mexican-American T2D = 174 Controls = 676 M and F PB 450 K TXNIP, ABCG1, SAMD12 associated with T2D, FBG, and HOMA-IR.
Soriano-Tarraga et al. (86) Caucasian, (Spain) aT2D = 151 Controls = 204 bT2D = 59 (BISMAR cohort) Controls = 108 bT2D = 63 (REGICOR cohort) Controls = 582 M and F WB 450 K One differentially methylated region in the TXNIP gene, replicated in 2 independent cohorts.
Florath et al. (87) German aT2D = 154 Controls = 835 bT2D = 87 Controls = 527 M and F WB 450 K Differential methylation of TXNIP associated with T2D in discovery and replication cohorts.
Jeon et al. (88) Korean aHigh-glucose group- 8 T2D = 5 Controls = 13 bT2D = 220 Controls = 220 M and F PB a450 K bBisulfite Pyrosequencing aMSI2 hypomethylated by 11% in T2D cases and 7% in high glucose cases (p-value = 0.038). CXXC4 hypomethylated by 15% in T2D cases (p-value = 0.044), and 12.8% in high glucose cases (p-value = 0.033).bMSI2 hypomethylation significantly correlated with T2D.
Yuan et al. (89) European aT2D = 23 Controls = 31 bT2D = 42 Controls = 221 M and F WB aMeDIP-seq and b450 K Two DMS within a 2 kb region upstream of the transcriptional start site of the MALT1 gene on T2D subjects compared to controls.
Matsha et al. (90) South African, mixed ancestry. T2D = 3 Prediabetes = 3 Controls = 3 F PB MeDIP-seq 1,415 DMS in the promoter regions of T2D subjects compared to normoglycaemic controls. Genes associated with cell surface signaling, glucose transport, insulin signaling, pancreas development, and the immune system.
Pheiffer et al. (11) South African, mixed ancestry. T2D = 3 Prediabetes = 3 Controls = 3 F PB MeDIP-seq 3,081 DMS in T2D and prediabetic subjects occurred within non-promoter regions, including sites encoding miRNAs.
a

Discovery cohort;

b

Validation cohort; DMS, differentially methylated sites; F, female; FBG, fasting blood glucose; FI, fasting insulin; HbA1c, glycated hemoglobin A1c; HOMA-IR, homeostatic model assessment-insulin resistance; M, male; MeDIP-seq, Methylated DNA immunoprecipitation sequencing; PB, Peripheral blood; T2D, Type 2 Diabetes; WB, Whole blood; 450 K, Infinium Human-Methylation450 BeadChip.